Help Fixing ZFS share

[arashi0987]

I was working on migrating my 4 drive ZFS Pool from my old Truenas Scale server to a Truenas Scale VM on Proxmox.
The setup is really like this: I have two SSDs on my computer, one with proxmox and one with Truenas Scale, and I boot off of one or the other throughout the process, so the drives don't really have to move
I exported the Pool (which is called Atlantic) on the Truenas Scale server And when I started up Proxmox, I was able to see Atlantic and import it.
So far so good.
Then I created a new VM of Truenas Scale, and passed through all of the Atlantic drives to the VM:


and on that VM it wanted to import the pool, so I let it, and I could see everything in Atlantic and it worked normally. However, the next time I rebooted proxmox, it would no longer mount the Atlantic ZFS Pool drives, and because my TrueNAS VM had those drives as passthrough drives. it wouldn't start the VM, so I couldn't mount the ZFS Pool in proxmox, and I couldn't launch TrueNAS to try to access it from there.

At this point I figured I might be able to try to reboot into my old TrueNAS Scale installation and import Atlantic there, but when I tried I got this error:


now when I try to import it back in proxmox, it sees the pool ,but it shows as degraded


If someone could help with mounting the pool, or in a worst case scenario, data recovery, I would appreciate it. I have a lot of data that would be hard to lose in this pool
Thanks

 

[arashi0987]

I've seen my mistake, I think. When I created the Truenas Scale Instance, I thought it created the scsi0 on my ssd that proxmox was installed on, but instead it put it on Atlantic, which I had intended to be owned by the Truenas VM.
So I think the problem started when the Truenas VM claimed it, but then was never able to be booted again. Does anyone know if there is a way to set the pool to ne able to be imported without being able to boot into the same VM that claimed the Pool?

 

[arashi0987]

Ok for anyone that has this problem in the future, my method of recovering the data was through installing a windows vm and using the Klennet ZFS recovery tool. It costs more than some might be able to afford, but the maker of the tool is very helpful, and will help in the recovery process.

 

[Dunuin]

However, the next time I rebooted proxmox, it would no longer mount the Atlantic ZFS Pool drives, and because my TrueNAS VM had those drives as passthrough drives. it wouldn't start the VM, so I couldn't mount the ZFS Pool in proxmox, and I couldn't launch TrueNAS to try to access it from there.

You don't want to import that pool or mount those disks on your PVE server when you want to passthrough them into your VM.
Passing through disks that are already mounted on the PVE host will corrupt its data.

I have a lot of data that would be hard to lose in this pool

Thats why you should have backups. Raid or snapshots don't count as those.

Ok for anyone that has this problem in the future, my method of recovering the data was through installing a windows vm and using the Klennet ZFS recovery tool. It costs more than some might be able to afford, but the maker of the tool is very helpful, and will help in the recovery process.

Great that you were able to recover data.

 

[arashi0987]

Thats why you should have backups. Raid or snapshots don't count as those.

Yep, now all of the data is being copied to my new backup server. Lesson learned